Consider the problem of jobs waiting in a queue until sufficient memory becomes available for them to be loaded and executed. If the queue is a simple first-in-first-out structure, then only the job at the head of the queue may be considered for placement in memory. With a more complex queuing mechanism, it might be possible to examine the entire queue to choose the next job to be loaded and executed. Show how the latter discipline, even though more complex, might yield better throughput than the simple first-in-first-out strategy. What problem could the latter approach suffer from?

What will be an ideal response?

Quite simply, if a small partition is available, or if several small partitions are available,
and if the job at the front of the queue is too large for any of the available partitions,
then small jobs in the queue behind the large job will be forced to wait. These small jobs
could be executing if the job-initiation mechanism could select them from their arbitrary
positions in the queue rather than from just the first position. The latter approach could suffer
from indefinite postponement of large jobs as smaller jobs continue arriving.